Even as many developed countries strengthen their traditional clinically-based surveillance capacity, a basic level of health information infrastructure is still lacking in many parts of the developing world, including areas that are most vulnerable to emerging health threats. To help fill this gap, electronic news outlets, disease reporting networks, discussion sites, and blogs are proving to be invaluable data sources for a new generation of public health surveillance systems that operate across international borders- almost all major outbreaks investigated by WHO, including SARS, are first identified by these informal online sources. These data sources provide local and timely information about disease outbreaks and related events around the world, including areas relatively invisible to day-to-day global public health efforts due to lack of infrastructure or political suppression of information. However, since this information is dispersed and largely unstructured, there is lack of organization and integration between sources, precluding a global view of all ongoing disease threats. In order to construct such an integrated view of current emerging infections, we deployed an early prototype called Health Map, a freely available multi-stream real-time knowledge management system that aggregates and maps health alerts across numerous key data sources, including WHO alerts, newswires, mailing lists. Leveraging our previous National Library of Medicine funded work in real-time health monitoring;we propose extensive development of this new digital resource to address technological and methodological barriers to achieving a synthesized, comprehensive, and customized view of global health. The principal objective of Health Map development is to provide access to the greatest amount of possibly useful information across the widest variety of geographical areas and disease agents, without overwhelming the user with an excess of information, or obscuring important and urgent elements. We will specifically address the challenge of providing structure to unstructured data while still enabling the astute user to notice the subtle pattern that could be an early indication of a significant outbreak. Development of Health Map will proceed along the four key components of surveillance: (1) data acquisition (evaluation and integration of multiple electronic sources) (2) information characterization (disease and location categorization and severity rating of unstructured data), (3) signal interpretation (multi-stream signal detection, spatiotemporal modeling and risk assessment) and (4) knowledge dissemination (tiered dissemination of global infectious disease alerts and flexible data visualization tools). In each area, we will make critical enhancements to the existing prototype. We plan rigorous evaluation to ensure that Health Map successfully leverages electronic sources for surveillance, communication, and intervention. We will also conduct comprehensive user testing, usability studies, user behavior analysis as part of our effort. Our goal is to make Health Map a vital knowledge resource tailored for both the general public and public health decision-makers.